Adobe offers Enhance Speech, a FREE online AI-powered speech enhancement tool.
I know that a free Creative Cloud membership gives you access to some apps and perks, but when I think of Adobe, I tend to go straight to the rather costly monthly subscription fees (for flagship products).
So, (fairly or unfairly) Adobe isn’t the first name that comes to mind when I think of developers who create free audio editing tools, but this one certainly looks impressive.
You can access the AI-powered speech enhancement tool via the Adobe Podcast website, which is free to join. To access the free tool, simply sign up for an Adobe user account.
Some online tools aim at live streams and balancing audio in real-time, but Adobe’s Speech Enhancement is about processing recorded audio.
You can process your audio by uploading your file or with a simple drag and drop, and that’s where your involvement ends. The process is entirely automatic and tackles a range of common problems.
I’m sure any of you who podcast or stream would be able to confirm that perfect recording conditions aren’t always easy to come by. Even when recording from a typically peaceful home studio, it’s not uncommon for construction workers to show up outside on the day with the noisiest machinery they can find.
No amount of planning can account for the random aspects of daily life that are out of our control, and by some cruel coincidence, it seems background noise will always find you.
Background noise is one of the biggest issues that Adobe’s free tool deals with, but it also tackles plosives, sibilance, and echo and might just make a cheap microphone sound like something much better.
As with any AI-powered tool, I can’t say the results will be perfect because there’s always the risk of harsh fixes. For example, AI technology might altogether remove unwanted artifacts but do so with a sledgehammer approach that leaves a word or two sounding harsh, abrupt, or weak.
However, what I’ve heard from this particular tool so far sounds great. The audio definitely sounds clearer, cleaner, and much more polished. It’s ideal for anyone who doesn’t have the tools or time to edit manually. I’d imagine that the ability to get consistent results so quickly for regular content creators could be a game-changer.
The video demonstrates the effectiveness of the AI speech enhancer in improving speech, however, its application to singing vocals is yet to be determined.
All I need now is an AI-powered tool that makes me sound exactly like Joe Pesci, Patrick Stewart, or Morgan Freeman (Editor’s Note: Better late than never, James!).
On a completely unrelated note, Spitfire Audio recently released a free piano library for LABS called Autograph Grand.
What is your viewpoint on the use of AI software in music production?
More info: Enhance Speech (requires free Adobe user account)
More:
16 Comments
Sir Otter
onIt’s great for de-noise and de-verb, otherwise you get better results manually if you know what you’re doing.
Tomislav Zlatic
onI tested it on a YT video I published recently, and just as you say, it did an incredible job of removing noise and reverb. The voiceover was recorded in a room with lots of bad reverb using a crappy lav mic. It sounded 10x better with Enhance Speech. However, I noticed that it removed the plosive from several words (for example “plugin” became “lugin”) so some improvements are needed for sure. But the tone enhancements and noise removal are incredible.
TORLEY
onThis Adobe tool actually reconstructs your signal. If there’s a lot of noise, your voice might end up sound very different. So unlike traditional noise reducers, it is NOT just removing parts of the waveform.
Try uploading a music file into it — it generates pitched vocal chops and other melodic glitches useful for resampling.
Tomislav Zlatic
onHey Torley, always lovely to hear from you! :) And thanks for the great idea, I’ll have to try processing some drums with this.
Steve A
onThere is something that feels morally abhorrent about software that “constructs” music based on analysing the music of flesh-and-blood artists. Although technically amazing, is it any more than “Artificial Plagiarism”?
On hearing an AI song allegedly constructed “in the style of Nick Cave”, the erudite Mr Cave is reported to have said “this song is bull****, a grotesque mockery of what it is to be human, and, well, I don’t much like it”.
CoopMusic247
onIf you wanna sound like someone else you should try
https://app.uberduck.ai/speak#mode=tts-basic&voice=zwf
There’s a better new one but this one does pretty “interesting” stuff
Daniel Calado
on“All I need now is an AI-powered tool that makes me sound exactly like Joe Pesci, Patrick Stewart, or Morgan Freeman”. LOL consider asking the guys at Krotos for that.
CoopMusic247
onFind it at uberduck
Arshan
onCheck out KRISP AI while you guys are at it. Discord has it inbuilt
Phil Huston
onThe Adobe Creative Cloud Free has many AI tools for graphic designers and audio. Add ChatGPT for words and pretty soon we’ll all be out work as the barely capable become craftspersons of content via AI!
Animus
onTried it. Heavy-handed. Added too much low-frequency muddiness and cut too much highs. Not very impressed.
TB
onIt sounds interesting.. but I won’t get involved with Adobe for anything, free or not.
Nizam
onMashallah
Nizam
onHi
Nizam
onHello
Vakeel khan
onHii